Adding layers with GeoAI

GeoAI can find geodata online and add it to your project as a new layer. This page covers the full workflow — from asking for data to approving and undoing layer adds — and explains each layer type GeoAI can create.

How layer adds work

When you ask GeoAI to find or add data, it follows a consistent workflow:

1. Search — GeoAI queries one or more data catalogs (Geonorge, OpenStreetMap, ArcGIS Hub, and so on) based on your request.
2. Validate — before proposing a layer, GeoAI checks that the data source is reachable and returns valid data.
3. Propose — an approval card appears in the chat panel with the layer name, type, and source.
4. You approve or skip — click Add to add the layer, or Skip to dismiss it. The layer is never added without your explicit confirmation.
5. Undo — after a layer is added, an Undo banner appears at the top of the panel. Click it to remove the layer immediately if it is not what you wanted.

tip

You do not need to type "yes" or "confirm". Just click the Add button on the approval card.

Layer types GeoAI can add

WMS layers

WMS (Web Map Service) layers stream map imagery from an OGC-compliant server. GeoAI verifies the service is online and checks which layers are available before proposing one.

Common sources: Geonorge, NVE, NASA GIBS, EMODnet, national mapping agencies.

"Find flood zone data for Oslo from Geonorge" "Add NASA GIBS true-colour imagery"

WMTS layers

WMTS (Web Map Tile Service) layers use pre-rendered cached tiles, which load faster than WMS — particularly useful for basemaps and imagery backgrounds that cover large areas.

GeoAI calls checkWmtsService first to discover the available layers, tile matrix sets, and styles from the service's capabilities document, then proposes the most relevant layer.

Common sources: Kartverket (Norway), IGN (France), Ordnance Survey (UK), Geonorge INSPIRE services.

"Add the Kartverket topographic map as a WMTS layer" "Add a WMTS basemap from the Geonorge INSPIRE service"

XYZ tile layers

XYZ tile layers use a URL template with {x}, {y}, {z} placeholders — the same format used by OpenStreetMap, Google Maps, and most web-based tile providers. They are lightweight and load quickly at any zoom level.

GeoAI validates that the URL template contains the required placeholders before proposing the layer.

Common sources: OpenStreetMap, Stadia Maps (formerly Stamen), MapTiler, Thunderforest, custom tile servers.

"Add OpenStreetMap as a background tile layer" "Add a terrain basemap from MapTiler"

info

XYZ tile layers are imagery overlays only — they do not carry attribute data that can be queried or analyzed. For interactive vector features, ask GeoAI to add a GeoJSON or ArcGIS FeatureServer layer instead.

GeoJSON layers

GeoJSON layers contain vector features (points, lines, or polygons) with attribute data. They support full styling, filtering, and analysis within AugmentCity.

GeoAI can fetch GeoJSON from public URLs or build Overpass API queries to extract OpenStreetMap features. For external URLs, it verifies the data is reachable and counts features before proposing.

"Add OpenStreetMap schools within 5 km of the city centre" "Find EU administrative boundaries for NUTS 2 regions"

COG raster layers

Cloud-Optimised GeoTIFF (COG) layers display georeferenced raster imagery such as elevation models, land-use classification maps, and satellite-derived products.

"Add a land-use raster from Copernicus"

info

COG layers are stored in your project but visual rendering in the viewer requires a Titiler tile server to be configured by your platform administrator. Contact your administrator if COG layers are not rendering.

ArcGIS layers

GeoAI can search ArcGIS Hub for public datasets and connect to ArcGIS REST services. It supports both:

• FeatureServer — interactive vector features (points, lines, polygons) with attribute data
• MapServer — tiled imagery overlays

GeoAI browses the service to discover available layers and confirms the correct layer ID before proposing.

"Find building footprints from the Oslo municipality ArcGIS server" "Search ArcGIS Hub for zoning data in Rotterdam"

Cesium Ion layers

Cesium Ion is a cloud platform that hosts 3D tilesets — including 3D buildings, photogrammetry captures, and high-resolution terrain. GeoAI can add Ion assets directly to your project using their numeric asset ID.

For terrain assets, GeoAI sets the elevation flag automatically so the terrain integrates correctly with the 3D viewer.

"Add Cesium Ion asset 96188 as a 3D tileset" "Add the Cesium World Terrain layer"

warning

Cesium Ion layers require a valid Cesium Ion access token to be configured in your project settings. If the token is missing or expired, the layer will not load. Contact your project owner to set up the Cesium Ion token.

I3S Scene Layer

I3S (Indexed 3D Scene Layers) is an OGC standard for streaming large 3D datasets such as city-scale buildings and terrain meshes. I3S services are commonly hosted on ArcGIS Online and ArcGIS Enterprise.

GeoAI can add an I3S layer directly from a Scene Service URL. Paste the URL into the chat and ask GeoAI to add it, or describe the dataset you need and GeoAI will search for a suitable service.

"Add an I3S layer from https://tiles.arcgis.com/…/SceneServer" "Add the 3D building model for Oslo from ArcGIS Online as an I3S layer"

See I3S Scene Layers for details on URL format and optional geoid terrain correction.

Google 3D Tiles

GeoAI can add a Google Photorealistic 3D Tiles layer to your project. This streams a global photorealistic 3D model of the real world directly from Google Maps, giving you a high-resolution base for your planning overlays.

"Add Google 3D Tiles" "Load the Google Photorealistic 3D base layer"

warning

Adding Google 3D Tiles requires a Google Maps API key with the Map Tiles API enabled to be configured in your project settings. GeoAI will prompt you to confirm before adding the layer. If the key is missing or not configured, the layer will not load.

KML/KMZ layers

KML (Keyhole Markup Language) and KMZ (compressed KML) files contain geographic annotations, placemarks, polygons, and network links. They are commonly exported from Google Earth and other GIS tools.

GeoAI verifies that the URL is reachable and returns a valid KML document before proposing the layer.

Maximum supported file size: 100 MB.

"Add this KML file from the regional planning portal: https://example.com/zones.kml" "Load the Google Earth export from the project shared drive"

The approval card

Every layer add goes through an approval card before being applied to your project. The card shows:

• The layer name GeoAI has chosen
• The layer type (WMS, WMTS, XYZ, GeoJSON, and so on)
• The source URL or asset reference

Click Add to add the layer. Click Skip to dismiss it without adding anything.

If GeoAI proposes multiple layers in one response, each layer gets its own approval card. Approve the ones you want and skip the rest.

The undo banner

After every approved layer add, an Undo banner appears at the top of the GeoAI panel. Click it to remove the layer immediately.

The banner is only visible for a short time. If you miss it, you can still remove the layer from the Layers panel using the delete button on the layer item.

The community catalog

GeoAI maintains a community catalog of geodata sources that have been successfully used across the platform. When you search for data, GeoAI checks this catalog first — community-verified sources are more likely to be online and return the data you expect.

The catalog now records all layer types, including WMS, GeoJSON, ArcGIS, WMTS, XYZ, and KML sources. Each time a layer from a known source is added successfully, its usage count increases, making it easier for GeoAI to recommend it to other users in similar contexts.

Noise data from Norwegian sources

GeoAI has a dedicated noise data search that queries Norwegian environmental data sources. It always includes the Miljødirektoratet støy WMS as a primary source and combines it with relevant noise datasets from the Geonorge catalog.

What you can ask for

You can specify a noise source type or leave it open to retrieve all available noise data:

Request	What GeoAI searches for
"Find noise maps for Oslo"	All noise data (road, railway, air, industry)
"Show road noise data for Bergen"	Road traffic noise layers
"Add railway noise contours for this area"	Railway noise layers
"Find aviation noise maps near Gardermoen"	Air traffic noise layers
"Find industrial noise data for this port"	Industrial noise layers

Noise metrics

GeoAI understands the standard Norwegian and EU noise metrics used in these datasets:

Metric	Description
Lden	24-hour average weighted for evening (+5 dB) and night (+10 dB) sensitivity — the primary metric used in Norwegian planning
Lnight	Night-time average (22:00–07:00) — used for sleep disturbance assessment
Støysone	Noise zones — red zone (highest exposure) and yellow zone (moderate exposure) as defined in Norwegian planning regulations

Standard noise color styling

After adding a noise WMS layer, you can ask GeoAI to apply the standard EU/Norwegian dB color ramp. The ramp follows the seven-step scale used in Norwegian environmental reporting:

dB level	Color	Meaning
≤ 45 dB	Green	Below significant impact threshold
45–50 dB	Light green	Low impact
50–55 dB	Yellow-green	Moderate impact
55–60 dB	Yellow	Action threshold area
60–65 dB	Orange	High impact
65–70 dB	Red	Støysone red zone (Norwegian standard)
≥ 70–75 dB	Dark red / purple	Severe impact

Ask GeoAI to apply this styling after the layer is added:

"Apply the standard noise color ramp to this layer" "Style the noise layer using the EU dB color scale"

info

Noise WMS layers use server-side rendering — the color styling is controlled by the WMS service itself and may not match the EU color ramp exactly. Where GeoAI applies client-side styling (for example on GeoJSON noise contour layers), the seven-step color ramp is applied automatically.

Tips for finding good data

• Include the country or region in your request: "flood zones in the Bergen municipality" gives better results than "flood zones".
• Include the data type you need: basemap, boundary, hazard, land use, population, and so on.
• If the first result is not right, ask GeoAI to try again: "That layer didn't have what I needed — try a different source".
• If you already know the URL of a WMS, WMTS, or KML service, paste it into the chat and ask GeoAI to add it. It will validate and propose the appropriate layer automatically.
• Use the wrench icon in the panel header to see tool call details and check exactly which URLs GeoAI validated or queried.